agent memory architecture

4 articles · 12 co-occurring · 0 contradictions · 0 briefs

Directly discusses where agent memory should live in system architecture—central to how context is managed across agent sessions.

Related concepts

tool integration patterns 4 multi agent orchestration 4 memory persistence 3 state management 2 state persistence across sessions 1 provider lock in vs portability 1 prompt engineering vs context engineering 1 observability as context validation 1 information flow management 1 guardrails and alignment 1 context window management 1 action space constraints 1

Evidence chain (4 articles, showing 4)

Engineering Trustworthy Multi-Agent Systems: A Deep Dive example_of

Tutorial explicitly discusses 'different types of short-term and long-term memory and how to utilize them for agents to follow business rules'—this is a concrete instance of context engineering applie

@hwchase17: > which is exactly why we believe memory should live outside of model providers example_of

Directly discusses where agent memory should live in system architecture—central to how context is managed across agent sessions.

Beyond AI Agents: Designing Agentic AI with Memory, Context, and Collaboration | by Vimal Dwarampudi | Medium example_of

Describes short-term and long-term memory modules as discrete components in step 4 of workflow, treating memory as separable from agent logic

Agents - CrewAI example_of

CrewAI's agent memory system is a concrete implementation of multi-turn context preservation across task sequences

query this concept

$ db.articles("agent-memory-architecture")

$ db.cooccurrence("agent-memory-architecture")

$ db.contradictions("agent-memory-architecture")